The author(s) shown below used Federal funds provided by the U.S. Department of Justice and prepared the following final report: Document Title: The Effect of Father Absence and Father
Alternatives on Female and Male Rates of Violence
Author(s): Jennifer Schwartz Document No.: 206316 Date Received: July 2004 Award Number: 2002-IJ-CX-0025 This report has not been published by the U.S. Department of Justice. To provide better customer service, NCJRS has made this Federally-funded grant final report available electronically in addition to traditional paper copies.
Opinions or points of view expressed are those
of the author(s) and do not necessarily reflect the official position or policies of the U.S.
Department of Justice.
The Pennsylvania State University
The Graduate School
College of the Liberal Arts
THE EFFECT OF FATHER ABSENCE AND FATHER ALTERNATIVES ON
FEMALE AND MALE RATES OF VIOLENCE
A Thesis in
Sociology
by
Jennifer Schwartz
copyright 2003 Jennifer Schwartz
Submitted in Partial Fulfillment of the Requirements
for the Degree of
Doctor of Philosophy
August 2003
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
Supported Under Award # 1 ) . from the National Institute of Justice Programs. Points of view in this document are those of the authors and do not necessarily represents the official position of the US. Department of Justice.
/72*;The disappearance of family-centered males from commu ties has generated considerable concern among policymakers and criminologists alike since past research on
the family structure of a community has revealed father absence to be a consistent and
potent predictor of variation in levels of violence across ecological contexts. However, it
is unclear why father absence is problematic (e.g., poor supervision, economic hardship,
lack of role models). Also at issue is whether alternatives to resident fathers and
husbands - such as older males referred to as "old heads" and co-resident grandparents -
can mitigate some of the negative effects of father absence on community levels of
violence. Attention to gender is particularly important to this issue of family structure
and community violence since, both historically and more recently, there is considerable
debate regarding whether family-related variables, and macro-structural predictors more
generally, are of more consequence to female or male offending or if they have a more
global effect.
Thus, the following research questions are posed: 1) Does father absence affect
female and male rates of violence similarly? 2) Do the effects of father absence persist
for both females and males once structural disadvantage and neighborhood guardianship
structures are taken into account? 3) Do alternatives to resident fathers, namely old heads
and co-resident grandparents, mitigate the negative effects of father absence on female
and male violence rates? and 4) To what extent do changes in father absence drive trends
in female and male levels of violence? County-level data from the UCR and the Census
Bureau are utilized with seemingly unrelated regression techniques and longitudinal
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
hierarchical models to address these issues.
Father absence has a strong and significant effect on both female and male levels
of violence across the three types of violence examined. Furthermore, the effects of
father absence are not differentiated by gender. Though both structural disadvantage and
supervisory structures influence community violence rates, the effects of father absence
persist even after taking these factors into consideration. The presence of other males in
the community and co-resident grandparents does mitigate some of the negative effects of
father absence, especially in areas where father absence is particularly acute. The
buffering effect of old heads is largely confined to homicide whereas the presence of co-
resident grandparents is effective across violent offenses. Results are not differentiated
by gender. Changes over time in the level of father absence in a community significantly
predict changes in female and male rates of violence, at least for homicide and robbery.
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
SUMMARIES
Over the past several decades, there has been a marked increase in the share of
families with absent fathers and husbands. For example, the percentage of children living
without a father more than tripled, from 12 percent in 1960 to almost 40 percent in 1995
(McLanahan 1997). This decline of resident fatherhood has generated considerable
concern among policymakers as much evidence indicates deleterious social consequences
of this trend. At the individual-level, children from father-absent homes tend to manifest
more behavioral problems including fighting and physical aggression (McLanahan 1997).
Further, the two-adult family form has been demonstrated to insulate its adult members
from involvement in violence as well (e.g., Wan: 1993).
Policymakers are concerned that the increase in non-traditional family forms is
indicative of a loss of community and social capital. The family institution has been
identified as a source of community functioning and stability that forms a barrier against
violence. Further, a community's family structure is indicative of its ability to effectively
socialize residents, exert informal social control, and garner social capital and resources
that buffer against violence (Biblarz and Raferty 1999). Compatible with this concern, at
the community-level, "family disorganization" has emerged across a number of
contemporary studies as one of the most potent and consistent predictors of community
variation in levels of adult (and juvenile) violence (Blau and Blau 1982; Sampson 1987;
Land, McCall, and Cohen 1990; Shihadeh and Steffensmeier 1994). It seems that the
scope of father absence helps to distinguish between communities in terms of level of
violent offending but also there is realistic concern that trends in father absence may have
important consequences for trends in violence.
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
Though it is clear from previous research that the family structure of a community
impacts its ability to resist violence, it is unclear exactly why widespread father absence
is problematic. Some scholars suggest that father absence is problematic because it
signals weak social control via poor supervision and decreased guardianship over public
areas, allowing interpersonal conflicts to escalate to criminal violence. Others assert that
father absence is merely a proxy for economic disadvantage and emphasize the lack of
social capital and resources in areas deficient in family-centered men. Additionally,
fathers collectively perform important socialization fbnctions within communities, acting
as role models and applying social stigma to enforce community norms. However, the
dynamics of the relationship between father absence, social control, structural
disadvantage, and socialization has rarely been examined at the community-level within
the criminological literature.
Given the important roles within communities that fathers play such that
pervasive father absence leads to high levels of violence, family sociologists and some
criminologists have suggested that the presence of other types of adults might together be
able to substitute for resident fathers and be able to mitigate some of the negative effects
of father absence. For example, Steffensmeier and Harer (1999) identify a relatively
recent shift in the nation's collective conscience, such that the aging cohort of baby
boomers exert a moral influence over communities as well as a greater willingness to
intervene for the common good. More specifically, Anderson (1990) identifies the
presence of "old heads" within a community as powerful agents of socialization and
social control for young adults. In fact Anderson even suggests that the old head "acted
as surrogate father to those who needed.. .moral support" (3). Additionally, the presence
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
of these hard-working, family-centered males might bring resources and social capital to
localities that would benefit the community as a whole. The female counterpart,
sometimes identified as "othermothers" or community mothers, is also integral to a
community's efforts to resist violence. These women, often grandmothers responsible
for child-care and general supervision over the area, are part of a network of collective
control and caregiving (Collins 2000). They are supportive of the family institution and
may alleviate some of the stress and strain experienced by single mothers. The presence
in communities of these alternatives to resident fathers may mitigate some of the negative
effects of father absence, yet until now this has not been empirically examined.
The approach of this study is unique in that it considers these issues from a
macro-perspective, with the community as the unit of analysis. Though macro-level
community research has witnessed a revival over the past decade, a focus on gender as an
analytical variable has remained virtually absent (but see Steffensmeier and Haynie
2000a, 2000b; Steffensmeier and Streifel 1992; Weisheit 1993). Attention to gender is
particularly poignant to the issue of family structure's impact on rates of violence as well
as to the larger debate regarding the universality of macro-structural predictors of
violence. Both historically and more recently there has been debate over whether family-
related variables are of more consequence to female offending or male offending. Many
have argued that females will be more affected by disruptions within the private, familial
sphere whereas males are more resistant to these pressures but more subject to economic
hardship (Broidy and Agnew 1997; Durkheim 1952; Elliot and Voss 1974; Messner
1985; Parmelee 191 8). Others contend that father absence will be more detrimental to
male levels of violence due to the lack of positive male role models and the presence of
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
an authority figure (Reiss 1986; Wilson 1987). Yet others claim the effects of non-
resident fathers are similar for males and females, arguing for global effects of father
absence and macro-structural predictors more generally (McLanahan 1985;
Steffensmeier and Haynie 2000).
Thus, there are several gaps in the criminological literature. The current research
links together the emergent focus on community-level family structure (and other
structural features of place) and the debate regarding the applicability of macro-factors to
female offending with the concern among criminologists and family sociologists
regarding the importance of family variables for female and male violence. Using data
from the Uniform Crime Reports and the 2000 Census aggregated to the county-level, we
explore the effects of father absence and alternatives to resident fathers on gender-
disaggregated rates of violence across counties. This study also seeks to understand
under what conditions father absence will have a criminogenic impact on communities.
In addition, we use an HLM longitudinal framework and data from earlier time points
(1 970-2000) to examine the dynamic relationship between trends in father absence and
female and male violent offending. Specifically, we ask: 1) Does father absence affect
female and male rates of violence similarly?; 2) Do the effects of father absence persist
for both males and females once community social control structures and structural
disadvantage are taken into account; and 3) Can alternatives to resident fathers such as
older responsible males (e.g., old heads) and care-taking grandmothers mitigate the
negative effects of father absence on violence for both males and females? After
establishing a firmer understanding of the functioning of father absence and gender-
disaggregated violent offending, we examine the dynamic nature of this relationship: To
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
what extent do changes in father absence drive trends in female and male levels of
violence?
This research goes beyond previous investigations by a) gender disaggregating
violent crime rates and indicator variables, where theoretically appropriate, across
metropolitan counties; b) drawing on a comparative approach to evaluate the effects of
father absence and other structural variables on both female and male rates of violence; c)
disentangling the family disruption-crime link by examining whether economic
disadvantage associated with father absence or the lack of supervision attributed to
female headed households is problematic; d) investigating whether alternate sources of
economic resources, social control, and socialization aside from resident fathers are
capable of mitigating female and male violence ;e) employing new analytic methods
(i.e., HLM growth modeling techniques) to examine how female and male violence
varies over time and place as a function of father absence.
Because this project examines the conditions under which father absence
compromises a community's ability to resist male and female violence, this project
requires ecological rather than individual-level data. Therefore, the unit of analysis for
the gender-disaggregated analyses of the impact of father absence on violent crime is the
approximately 2000 counties on which violent crime data are available for 2000. The
dependent variables, adult female and male violent crime rates, are obtained fiom the
Uniform Crime Reporting Program for police agencies in metropolitan areas. The
agency-level crime data has been corrected for incomplete reporting by police agencies
(i.e., less than 12 months of data); aggregated to the county level; adjusted for incomplete
population coverage (i.e., non-participating or missing police agencies); averaged over a
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
three-year period to avoid annual fluctuations in events as rare as violent crime; and
transformed into crime rates using US Census Bureau population figures to yield sex-
specific county level arrest rates for those 18 and over. The sex-specific dependent
variables are logarithmically transformed to induce homoscedasticity and to counteract
the possible floor effect of these right-skewed distributions. Violent crimes included in
this analysis are homicide, robbery and felony assault.
The data for the independent and control variables are derived from Bureau of
Census Summary Tape Files 3 and 4 (STF). Key variables include father absence
(defined as the percent of families that are headed by females), male capital (presence of
old heads), community caregivers (presence of caregiving grandparents), structural
disadvantage, supervisory structure (measure of available guardians), and a number of
control variables (e.g., age structure, residential stability, etc.).
"Seemingly unrelated regression" techniques (SUR) are used to estimate separate
models for females and males. This method corrects for correlated error across the
equations. F-tests for equality of coefficients between female and male models estimated
from this approach are used to determine if there are any significant differences across
sub-groups in the strength of the effects of the independent and control variables. An
overdispersed, hierarchical linear model is used to conduct a growth-curve analysis of the
impact of changing family structure on within-county trends in female and male violence.
There were several key findings of this research. Father absence had strong and
significant effects on variation in female and male violence rates across ecological
contexts. Further, the effects of father absence are largely similar in magnitude for
females and males. Both cross-sectional seemingly unrelated regression models from
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
1990 and 2000 as well as longitudinal growth curve models estimated using hierarchical
linear modeling support these conclusions. Father absence as a predictor of violence is
extremely robust for both female and male violence. Even after controlling for factors
related to the deleterious consequences of father absence - structural disadvantage and
community social control mechanisms- the association between father absence and
female and male offending remained. Though these two factors likely mediate some of
the effect of father absence on violence, other processes that are not as amenable to
measurement using Census indicators, such as the capacity of resident fathers to mentor,
protect, and act as moral compasses within communities, are likely at work.
Though the absence of resident fathers continues to exert direct effects on female
and male violence, the presence of male capital and collective caregiving within
communities can mitigate some of the violence producing effects of father absence. It is
likely that the increased presence of a relatively large cohort of older males has a direct
violence-reducing effect on both female and male violence, though results were not
definitive. Old heads were found to be more effective in curbing homicide in localities
that acutely lacked resident fathers, though they did little to stem robbery or felony
assault and may even work to increase these types of violent offending by females and
males. Results did not vary by gender. Though collective caregiving by resident
grandparents was not found to be an effective alternative to resident fathers on average,
in high father absence areas, the increased presence of these caregivers was significantly
associated with reduced rates of violence across gender and violence type. Thus, the
presence of grandmothers has a clear buffering effect where father absence is particularly
acute for both females and males. However, it should be noted that although male capital
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
and collective care-giving made up for some of the community deficits associated with
widespread father absence, father absence continued to exert significant, violence
producing effects on gender-disaggregated violence rates.
Consistent with the robust nature of the cross-sectional father absence-violence
relationship for both females and males, longitudinal models suggest changes in father
absence are significantly related to variability in trends in homicide and robbery, but not
felony assault. Further, trends in father absence explained a substantial proportion of
variance (i.e., overdispersion) in trends in violence. Few significant gender differences
emerged in the relationship between changes in father absence and trends in violence.
Though this project represents a considerable advance in the field's knowledge of
the relationship between family structure and gender-disaggregated rates of violence
across ecological contexts, several questions emerge and there are a number of caveats of
the current study to be attended to by future research. The main issues discussed include:
Why might we expect such similarities across gender in the effect of father absence on
violence across ecological contexts? In what ways and why might older adults benefit
communities in terms of violent crime? In light of the mixed findings regarding male
capital, why might older males be detrimental to community violence control efforts?
Related to these issues, are there gender differences in the ability of older adults to
suppress violence within communities? In reviewing and exploring the nature of findings
from this research, a number of caveats and directions for future research are embedded
throughout the discussion. They include: the paucity of official statistics as measures of
both violence and independent variables; the failure of this research to explore linkages
between micro- and macro- factors associated with the family structure-violence
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
relationship; and the inability of this research to address issues of race and ethnicity.
Future research should examine whether these relationships vary by gender*ethnicity; use
alternate sources of data to verify these findings; examine whether the institution of old
heads has changed over time; and qualitatively flesh out the mediating mechanisms of
identified relationships between gender, father absence, and violence.
The findings from this research suggest that solutions to violence for both females
and males can be applied at the community-level. Of note, the marked similarities in
factors driving community violence levels for both females and males suggests that
prescriptions against violence are, to some extent, gender neutral. Female and male
violence are entwined so measures taken to reduce male violence are also likely to reduce
female violence, and vice versa. By the same token, there were also some gender
differences identified and it was suggested that the mechanisms underlying female and
male violence might differ. As such, gender-specific policies that identify what it means
to be "female" or "male" are complementary to more global violence-reducing policies.
Regarding the finding that the presence of fathers strengthens community abilities
to resist violence, the policy implications are to make men more "marriageable" as
partners. This would require that there be sufficient job opportunities at the requisite
skill-level of residents. In addition, incarceration policies that remove males from the
community may have negative consequences for these areas in at least two respects.
First, the large-scale removal of young males from a community tips the sex ratio such
that competition over males is intensified. Research on family formation processes
suggests that marriage is less likely where males are scarce. Further, the competition
over marriageable partners is hypothesized to aggravate stress among women in a way
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
that fosters female violence. Second, the imprisonment of males, especially for non-
violent (e.g., drug) offenses stigmatizes a substantial portion of males in some
communities such that the ability to secure viable employment in the legitimate sector is
severely handicapped. Not only does this reduce the "marriageability" of males in these
types of areas, but the economic deprivation associated with unemployment also
exacerbates the likelihood of male (as well as female) violence. The findings from this
research also suggest that extended kin networks, primarily grandmothers, can alleviate
some of the violence-producing qualities of high father absence areas. Governmental
support to co-resident grandparents, including health care as well as transfer payments,
that recognizes the importance of these networks is warranted. Also of note, the findings
regarding the similarity of underlying factors that produce violence for females and males
suggests that community solutions
There are also some micro-level policy implications as well. The results from this
research suggest that, though the economics and supervisory capacity of an area are
important, there are other immeasurable qualities that fathers offer to communities, such
as mentorship. As such, Big BrotherISister programs and other programs that create
opportunities for mentorship may work to reduce violence. Supervision and guardianship
of communities is also important, so programs that engage youth in after-school activities
and young adults in activities such as work or legitimate leisure pursuits (e.g., the arts,
sports, etc.) might aid mother-only families in the ability to supervise young adults as
well as offer opportunities to create networks of collective control.
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
Chapter 1: Introduction and Literature Review
Introduction
Over the past several decades, there has been a marked increase in the share of
families with absent fathers and husbands. For example, the percentage of children living
without a father more than tripled, from 12 percent in 1960 to almost 40 percent in 1995
(McLanahan 1997). This decline of resident fatherhood has generated considerable
concern among policymakers as much evidence indicates deleterious social consequences
of this trend. At the individual-level, children from father-absent homes tend to manifest
more behavioral problems including fighting and physical aggression (McLanahan 1997).
Further, the two-adult family form has been demonstrated to insulate its adult members
from involvement in violence as well (e.g., Warr 1993).
Policymakers are concerned that the increase in non-traditional family forms is
indicative of a loss of community and social capital. The family institution has been
identified as a source of community functioning and stability that forms a barrier against
violence. Further, a community's family structure is indicative of its ability to effectively
socialize residents, exert informal social control, and garner social capital and resources
that buffer against violence (Biblarz and Raferty 1999). Compatible with this concern, at
the community-level, "family disorganization" has emerged across a number of
contemporary studies as one of the most potent and consistent predictors of community
variation in levels of adult (and juvenile) violence (Blau and Blau 1982; Sampson 1987;
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
Land, McCall, and Cohen 1990; Shihadeh and Steffensmeier 1994). It seems that the
scope of father absence helps to distinguish between communities in terms of level of
violent offending but also there is realistic concern that trends in father absence may have
important consequences for trends in violence.
Though it is clear from previous research that the family structure of a community
impacts its ability to resist violence, it is unclear exactly why widespread father absence
is problematic. Some scholars suggest that father absence is problematic because it
signals weak social control via poor supervision and decreased guardianship over public
areas, allowing interpersonal conflicts to escalate to criminal violence. Others assert that
father absence is merely a proxy for economic disadvantage and emphasize the lack of
social capital and resources in areas deficient in family-centered men. Additionally,
fathers collectively perform important socialization functions within communities, acting
as role models and applying social stigma to enforce community norms. However, the
dynamics of the relationship between father absence, social control, structural
disadvantage, and socialization has rarely been examined at the community-level within
the criminological literature.
Given the important roles within communities that fathers play such that
pervasive father absence leads to high levels of violence, family sociologists and some
criminologists have suggested that the presence of other types of adults might together be
able to substitute for resident fathers and be able to mitigate some of the negative effects
of father absence. For example, Steffensmeier and Harer (1999) identify a relatively
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
recent shift in the nation's collective conscience, such that the aging cohort of baby
boomers exert a moral influence over communities as well as a greater willingness to
intervene for the common good. More specifically, Anderson (1 990) identifies the
presence of "old heads" within a community as powerful agents of socialization and
social control for young adults. In fact Anderson even suggests that the old head "acted
as surrogate father to those who needed.. .moral support" (3). Additionally, the presence
of these hard-working, family-centered males might bring resources and social capital to
localities that would benefit the community as a whole. The female counterpart,
sometimes identified as "othermothers" or community mothers, is also integral to a
community's efforts to resist violence. These women, often grandmothers responsible
for child-care and general supervision over the area, are part of a network of collective
control and caregiving (Collins 2000). They are supportive of the family institution and
may alleviate some of the stress and strain experienced by single mothers. The presence
in communities of these alternatives to resident fathers may mitigate some of the negative
effects of father absence, yet until now this has not been empirically examined.
The approach of this dissertation study is unique in that it considers these issues
from a macro-perspective, with the community as the unit of analysis. Though macro-
level community research has witnessed a revival over the past decade, a focus on gender
as an analytical variable has remained virtually absent (but see Steffensmeier and Haynie
2000a, 2000b; Steffensmeier and Streifel 1992; Weisheit 1993). Attention to gender is
particularly poignant to the issue of family structure's impact on rates of violence as well
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
as to the larger debate regarding the universality of macro-structural predictors of
violence. Both historically and more recently there has been debate over whether family-
related variables are of more consequence to female offending or male offending. Many
have argued that females will be more affected by disruptions within the private, familial
sphere whereas males are more resistant to these pressures but more subject to economic
hardship (Broidy and Agnew 1997; Durkheim 1952; Elliot and Voss 1974; Messner
1985; Parmelee 19 18). Others contend that father absence will be more detrimental to
male levels of violence due to the lack of positive male role models and the presence of
an authority figure (Reiss 1986; Wilson 1987). Yet others claim the effects of non-
resident fathers are similar for males and females, arguing for global effects of father
absence and macro-structural predictors more generally (McLanahan 1985;
Steffensmeier and Haynie 2000).
Thus, there are several gaps in the criminological literature. The current research
links together the emergent focus on community-level family structure (and other
structural features of place) and the debate regarding the applicability of macro-factors to
female offending with the concern among criminologists and family sociologists
regarding the importance of family variables for female and male violence. Using data
from the Uniform Crime Reports and the 2000 Census aggregated to the county-level, we
explore the effects of father absence and alternatives to resident fathers on gender-
disaggregated rates of violence across counties. This study also seeks to understand
under what conditions father absence will have a criminogenic impact on communities.
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
In addition, we use an HLM longitudinal framework and data from earlier time points
(1970-2000) to examine the dynamic relationship between trends in father absence and
female and male violent offending. Specifically, we ask: 1) Does father absence affect
female and male rates of violence similarly?; 2) Do the effects of father absence persist
for both males and females once community social control structures and structural
disadvantage are taken into account; and 3) Can alternatives to resident fathers such as
older responsible males (e.g., old heads) and care-taking grandmothers mitigate the
negative effects of father absence on violence for both males and females? After
establishing a firmer understanding of the functioning of father absence and gender-
disaggregated violent offending, we examine the dynamic nature of this relationship: To
what extent do changes in father absence drive trends in female and male levels of
violence?
This research goes beyond previous investigations by a) gender disaggregating
violent crime rates and indicator variables, where theoretically appropriate, across
metropolitan counties; b) drawing on a comparative approach to evaluate the effects of
father absence and other structural variables on both female and male rates of violence; c)
disentangling the family disruption-crime link by examining whether economic
disadvantage associated with father absence or the lack of supervision attributed to
female headed households is problematic; d) investigating whether alternate sources of
economic resources, social control, and socialization aside from resident fathers are
capable of mitigating female and male violence ;e) employmg new analytic methods
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
(i.e., HLM growth modeling techniques) to examine how female and male violence
varies over time and place as a fbnction of father absence.
This chapter begins with a review of prior literature on the general relationship
between father absence and crime with specific emphasis on social control and
stratification explanations of the deleterious consequences of father absence for
communities. Next, the potential role of alternatives to resident fathers (e.g., old heads
and grandmothers) in violence prevention is discussed. The chapter concludes with a
rationale for expecting (or not expecting) gender differences in the impact of father
absence and other guardians on gender-disaggregated rates of violence.
Father Absence and Violent Crime
Concern with family disorganization as a source of delinquency dates back from
at least the 1 9th century (Bellingham 1986). For example W.I. Thomas (1 927)
emphasized the role of "broken homes" as a pathway to delinquency. A breakdown in
the family - the main socializing unit -may lead to inadequate socialization or bonding
which might result in norm breaking and law violation. Nye (1958) argued that family
structure impacted violence indirectly, via loss of direct parental control but also by
decreased parent-child attachments. However, this traditional premise of an individual-
level link between family disruption and crime has not received much empirical support.
Evidence of an individual-level link between family structure and crime is weak and
inconsistent, particularly for the more serious and violent crimes (Cernovich and
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
Giordano 1987; Ensminger, Kellam, and Rubin 1983; LaFree, Drass, and O'Day 1992;
Ross and Sawhill 1975; Shihadeh and Steffensmeier 1994; Wilkinson 1980).
In contrast to individualistic explanations, more current thinking on the
relationship between family structure and crime has built on elements of systemic social
disorganization theory and has consequently shifted in emphasis to the structural effects
of widespread father absence on community levels of violence. Though not hypothesized
to be a variable of consequence by Shaw and McKay, recent reformulations and
extensions of social disorganization theory have identified father absence as a potent
variable affecting comrnunity-level variations in violent offending. Indeed, macro-level
research consistently finds positive effects of family disruption1 on rates of violence
(Blau and Blau 1982; Huff-Corzine et a1 1986; Land, McCall, and Cohen 1990; Messner
1985; Messner and Golden 1992; Messner and Sampson 199 1 ;Ousey 1999; Sampson
1986,1987; Shihadeh and Steffensmeier 1994; Shihadeh and Flynn 1996; Simpson 1985;
Williams 1983; Williams and Flewelling 1988). In order to explain this relationship, two
main accounts have developed -one emphasizing the social control mechanisms that are
more prevalent where two-parent families predominate, the other emphasizing the
economic stratification and disadvantage elements ubiquitous among areas characterized
by many single-parent families.
Some scholars suggest the relationship between father absence and crime is due to
a community's diminished capacity to exercise informal social controls over its residents
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
(e.g., Messner and Sampson 199 1 ;Sampson 1987). Where father absence is common,
poor supervision of young adults and decreased guardianship over public areas may allow
verbal arguments to develop into aggressive acts of violence and criminal activity to
flourish (Sampson 1987). Others, however, emphasize the lack of human and social
capital in localities where father absence is prevalent (Rose and Clear 1998). These
scholars argue that communities suffer a net loss in economic and political power when
males are absent. Theorists suggest that economic distress is a precipitator of negative
social conditions that undermine legitimacy and weaken social norms restraining the use
of violence (Blau and Blau 1982). The milieu effects of deprivation-related fixstration
extend to all residents within the social context, regardless of their individual economic
circumstances or family structure. Though both the social control and stratification
explanations of the family disruption-crime link have received some empirical support,
the two have yet to be empirically disentangled2. Indeed, Sampson (1987, p. 376)
emphasizes that the mechanism linking family disruption with elevated levels of violence
"await[s] further research." Further, fathers likely perform collective functions in
addition to social control and economic provision, including roles as mentors andlor
protectors. We first examine how pervasive father absence impacts community-level
I In some of these studies, family disruption was operationalized as percent divorced or percent of children under 18 not living with both parents rather than percent female-headed families as operationalized here. 2 In fact, it is common to link the proportion of female headed households with measures of poverty and other indicators to create a "disadvantage index" (e.g., Land, McCall, and Cohen 1990).
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
social control and stratification3; we then speculate about other contributions that fathers
as a group make to communities.
Social Control: Father Absence and Lack of Supervision
Ecological theories of crime have typically been concerned with the impact of
social control on crime and, thus, have emphasized differences in the capacity of
communities to control violence. It is purported that a high fraction of single-parent
families in a community attenuate informal social controls that serve to restrain
threatening behaviors and criminal acts. Though the precise mechanisms by which father
absence contributes to violent crime have not been tested, systemic social disorganization
theorists argue that the control capacity of areas with high family disorganization is
weakened via decreased supervision; simply put, there are fewer guardians due to absent
fathers and husbands. In addition, localities with many female-headed households are
less capable of exerting informal control due to the time constraints on single mothers,
who tend to have higher rates of labor force participation (Waite 1981). Limited free
time of single working mothers may hinder supervisory behavior, organizational
participation, and contact with neighbors4 (Messner and Sampson, 1991 ;Sampson 1985,
The framework of communities as "units of stratification" and "units of social control" is drawn from Shihadeh and Steffensmeier 1994). 4 A good deal of empirical evidence suggests that community levels of family disruption do not alter the extensiveness of social networks/social interaction or organizational participation (Sampson and Groves 1989; Bellair 1997; Campbell and Lee 1992). However, though family disorganization does not impact the emergence of ties, it does seem to alter the effectiveness of those ties: "In short, social ties among women in communities with many female-headed households may not involve violence-preventing supervisory behavior to the same extent as ties in other neighborhoods. Women in such contexts.. .may still lack the resources necessary to extend such networks into the realm of supervision of potential offenders and
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
1986, 1987; Sampson and Groves 1989) - all of which are theorized to have preventive
effects on violence.
Two-parent households provide increased supervision over their own children and
property (Cohen and Felson 1979) but, perhaps more importantly, the opportunity for
deviant behavior is less where two-parent families are the norm due to the guardianship
exercised over the public activities of others within the community (Sampson 1987).
Guardianship includes acts such as taking note of andlor questioning strangers,
supervising youth activities and peer groups, watching over one another's property, and
intervening in local disturbances (Messner and Sampson 1991). More importantly than
intervening in actual criminal acts, guardians serve a preventive function by which they
are better able to control the activities of peer groups (e.g., "hanging out," truancy,
vandalism) that set the context for more serious involvement in violence and gang
delinquency by adults (Sampson 1987). In addition, unsupervised peer activity has been
demonstrated to contribute to higher levels of deviance and violence (Sampson and
Groves 1989).
Thus, the presence of resident fathers serves an important social control function
within communities by supervising public activities within the community,
supplementing female authority, and intervening into conflicts before they get out of
hand. This "community guardianship role" is socially expected of family-centered men
more so than women or unattached, single men. However, aside from the absence or
intervention in violence" (Rountree and Warner 1999: 806-807). As suggested below, either male authority
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
presence of family centered men, localities differ in their capacities to exert informal
social control. Perhaps areas that are rich in community guardians and supervisory
structures (e.g., as measured by the presence of available guardians) can overcome
deficits in resident fathers. If the primary function of fathers is related to social control,
the addition of direct measures of community supervisory structures should erase the
effects of father absence on violence. Therefore, we ask: Controlling for social control
mechanisms within the community, does the effect of father absence on female and male
violence persist?
Social Stratification: Father Absence and Lack of Community Resources
A common theme of communities and crime research is the importance of
economic hardship for influencing rates of violence. It was empirically documented as
early as the turn of the 2oth century that crime tends to be concentrated where resource
deprivation in concentrated (Shaw and McKay 1942), and more recent empirical work
confirms the continued importance of economic marginalization in elevating male andlor
total rates of violent crime (Blau and Blau 1982; Land, McCall, and Cohen 1990;
Patterson 199 1 ;Williams and Flewelling 1988). Theorists suggest that economic distress
is a precipitator of negative social conditions that undermine legitimacy and consequently
weaken social control and norms restraining the use of violence (Blau and Blau 1982).
The milieu effects of frustration related to feelings of deprivation extend to all residents
or economic resources may account for these conclusions.
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
within the social context, regardless of their individual economic position/family
structure (Allan and Steffensmeier 1989).
As Wilson (1 987; 1993) empirically demonstrates, father-absent households are
far more vulnerable to poverty than other family types, with these types of households
having poverty rates almost five times that of married-couple families. In addition, non-
resident father families are more likely than father-headed families to be persistently
poor. Moreover, female-headed households typically have less earning potential than
other family forms and less labor force attachmentlactivity (Wilson 1996: 93-94). Not
only is there an association between economic deprivation and family form at the
individual level, but female headed households are more likely to be found in areas of
concentrated poverty (see Appendix A ) . ~ This close association between family form and
living in circumstances of poverty has continued to plague researchers. A common
approach to dealing with the close association between poverty and family disruption has
been to create an index of these components (e.g., Land, McCall, and Cohen 1990).
Typically, indicators of economic disadvantage (e.g., median income, percent in poverty,
inequality) load on the same factor as family form. However, as Figuera-McDonough
(1 992) points out, on a theoretical level, researchers are confusing an organizational
element of communities (family disruption) with a demographic feature of communities
(feminization of poverty/resource deprivation).
This being the case, it is also important to note that there is considerable variation in levels of father absence and poverty across counties.
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
It is important to note that research addressing the impact of residing in a female-
headed family indicates that features of the area, such as poverty, may contextualize the
negative effects of these types of households. For example, families headed by females
are only vulnerable if they lack resources (Hogan and Kitagawa 1985; Cernovich and
Giordano 1987; Matsueda and Heimer 1987). These results suggest that it is not (solely)
the lack of supervision/control exercised in areas of extensive father absence that allows
crime to occur. Rather, the lack of economic resources found in areas with many absent
fathers may create contexts conducive to crime, as theorized above.
Men, especially men in families, provide many economic resources to a
community. For example, men in families typically provide monetary support, which
may allow women's social networks to function more adequately to control crime; the
added resources available in two-parent homes allows women to devote more energy to
the community networks that have been found to deter crime (Rountree and Warner
1999). Moreover, because men tend to have broader ties than women, males provide
extensive linkages to resources external to the community, including those that interact
with formal agents of social control and other sources of assistance to the community.
Thus, the absence of family-centered males can have a significant impact on the
social capital and availability of financial resources a community can draw upon. This
economic disadvantage, in turn, can create contexts that are relatively more conducive to
adult criminal violence via the effect of deprivation on frustration that leads to acts of
violence and on norms that restraidcontrol the use of violence. But, because most
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
research has combined father absence and resource deprivation, it has been difficult to
disentangle the relationship between family form and violence6. Therefore, this research
asks: Do father absence effects persist after structural disadvantage is taken into
account?
Other (Immeasurable) Functions of Resident Fathers: Socialization, Protection, and
Stabilitv
As alluded to earlier, the role of fathers within communities is much more
nuanced and involved than mere agents of social control and providers of resources and
social capital. Fathers also fill roles involving socialization (e.g., role modeling,
developing empathy, teaching gender roles) and protection (e-g., daughters from abuse,
sons from violence, and wives from assault). Further, the dual-parent family form is
mutually beneficial for adult men and women and provides stability to the community as
a whole. All of these factors create conditions that are incompatible with violence when
fathers are largely present in the community.
Fathers as agents of socialization act as role models for both boys and girls within
the community and offer guidance and instruction in ways that restrain the use of
violence. For males, fathers as a group teach boys about male responsibility and how to
There are several studies that examine family form as separate from economic indicators, though none of these studies focused specifically on the relationship between the two. Sampson (1987) finds significant positive effects of female headed households on white and black robbery offending rates, controlling for per capita income and welfare payments as indicators of disadvantage; per capita income generally has null effects. Shihadeh and Steffensmeier (1994) examined the indirect effect of inequality on violent crime. They find that income inequality increases family disruption, which elevates rates of black violence. Lastly, Blau and Golden (1986) find that criminal violence is higher in metropolitan areas where family
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
- -- - --
be suitably assertive (Popenoe 1996). Further, fathers prevent male peer groups fiom
becoming too dominant a force within communities (Anderson 1990). For females,
fathers teach daughters how to trust and relate to males and socialize them to choose
"marriageable" partners (i.e., those not involved in violence) (Popenoe 1996).
Crucial for the control of violence, empathy and messages unfavorable to the
commission of violence must be disseminated (Gilligan 1982). Developmental
psychologists assert that fathers are more important for the development of socially
acceptable forms of behavior that do not include violence and aggression (Popenoe
1996). In other words, fathers teach self-control and are models for compassion and
empathy, for both males and females, in ways that quell violence. Fathers are important
sources of messages unfavorable to interpersonal violence that permeate throughout the
community.
Resident fathers impose expectations about gender roles that serve to prevent
criminal violence whereas communities that are deficient in resident fathers lack an
important source of socialization for young men and women. For males, absent fathers
and the lack of male role models lead to greater interpersonal violence and exploitative
behavior toward women (Draper and Harpending 1982). Further, in the absence of
responsible, mature males, young men develop "hypermasculine" behavior in an attempt
to prove what it means to be a man through violent behavior (Popenoe 1996: 156). For
females, violent behavior is inconsistent with traditional gender expectations that
disruption (as measured by % divorced/separated) is greater; in addition, these rates are unrelated to an
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
encourage women to be soft and yielding (Steffensmeier and Allan 1996); to the extent
that females are discouraged from violent responses and other "unfeminine" behavior by
husbands and fathers, there will be less female violence in the community. Further,
where fathers and husbands are largely present, females are more insulated from
involvement in the breaking up of fights, surveillance, and street culture in general and its
concomitant violence. Also, females from father absent areas are likely to have more
relations with males at a younger age with fewer restrictions regarding sexual
involvement; this leads to greater involvement in violence because female violence is
often a response to provocation from males or a result of co-offending with males (Miller
1986).
Areas with an adequate supply of fathers offer protection to daughters, sons, and
wives, with beneficial effects regardless of gender. Daughters and wives are protected
from abuse from young males and strangers given the authority of mature, responsible,
family-oriented males. Also, as stated above, males are more restrained in their use of
violence where family men are available to socialize young males to acceptable
thresholds of aggression and violence. Sons are protected from violence via regulation of
the dominance of the peer group and the social expectation that fathers will step in to
break up fights (Popenoe 1996). Popenoe states "Neighborhoods without fathers are
neighborhoods without men able and willing to confront errant youth, chase threatening
gangs, and reproach delinquent fathers" (1 996: 140).
area's poverty level once other conditions are controlled.
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
Finally, areas that are plentiful in dual-parent homes enjoy more overall stability,
which is beneficial to communities as a whole (Durkheim 1952). Resident fatherhood is
a mutually beneficial family form for both adults. It prevents violence via the mutual
support offered to men and women in stable relationships that keeps stress and strain at a
minimum; provides stakes in conformity by offering strong disincentives to engage in
criminal violence; and brings access to broader and more differentiated social networks
of relatives and fhends that provide social capital and more economic resources as well
as offering coping mechanisms and sources of social support. Further, greater father
presence assists the community by freeing mothers to participate in community
organizations and build linkages and social capital that ultimately act to restrict violence
within the community.
Given the important and varied roles that cooperatives of fathers play in terms of
community violence prevention, it is vital that we understand the conditions under which
father absence is least detrimental. Though some argue that the importance of biological
fathers cannot be underestimated (e.g., Popenoe, 1996), others suggest that communities
can draw on alternative sources of resident fathers - including older responsible males
and involved, caregiving grandmothers.
Community Alternatives to Resident Fathers: Male Capital (Old Heads) and Community Caregivers (Grandmothers)
Studies of macro-level links between family form and violent crime typically do
not take into account the community's capacity for alternative forms of social control,
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
social capital, and socialization aside from two-parent families7. The proportion of older
adult males andlor the share of care-taker grandparents at large in the community are part
of a network of collective control. These older adults can play vital roles as role models,
guardians, or mediators who can have a significant impact on the prohibition of violence.
First we discuss the role of male capital -what the presence of a surplus of older men can
bring to a community; then we examine the potential impact of grandmothers as
community caregivers. Throughout this section, we highlight the importance of gender in
both preventing and engaging in acts of violence.
Male Capital and Old Heads
The presence of older, responsible males in an area has been identified as a
central aspect of social organization within communities. Anderson (1990) describes
members of this community institution as hardworking men who aggressively sought out
young men in their late teens or early twenties in order to socialize them to fulfill their
responsibilities regarding work, family, and the law and to assist them in their transition
to responsible adult roles. In addition to preaching an anti-crime and anti-trouble
message, old heads acted as guidance counselors, advising young adults and intervening
in trouble (1990: 69) - shepherding these young people into productive adulthood. These
Though Messner and Sampson (1 99 1) explore the effects of male presence on violent crime as mediated though family disruption, their intent is not to parcel out alternative guardianship practices in a community. The focus, rather, is on the influence of a shortage of males on family formation processes and the subsequent influence of female headed households on violence. Thus, their measure of sex composition is the number of males per 100 females ages 15-59. They conclude that an area's sex ratio impacts violent crime only through family formation processes, which directly influence violent crime. Hence, there is evidence that the demographic makeup of a community can have a significant impact on its violent crime rate, albeit indirectly.
7
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
older men, typically in their fifties or sixties, often "acted as a surrogate father to those
who needed attention, care, and moral support" (Anderson 1990:3). They were
considered role models within the community as well as authority figures. This
highlights the potential importance of their role in communities that lack resident fathers
as they perform many of the social functions usually filled by biological fathers. The old
head also provides a measure of social capital to a community. He represents an
important linkage between the community and larger society, both through his own
legitimate employment and also via finding paid employment for young men outside the
boundaries of the community (1990:71). The old head's role of surrogate father, role
model, authority figure, and provider of social capital and the implications for effective
crime control are now discussed in more detail.
Anderson (1990) and Wilson (1987) emphasize the importance of adult male
networks within communities in order for effective crime control to take place. They
argue that in contexts of widespread family disruption, the effectiveness of informal
social control in preventing violence can break down. The authority of women,
particularly in the context of single-parent families, may carry less weight when not
supported by male authority. In particular, mother-only families are especially
vulnerable to the challenge of male peer groups: "In those domestic situations in which
there is only one adult.. .the dwelling may be viewed by young boys, superficially at
least, as essentially an unprotected nest.. .Insuch a setting, a man, the figure the boys are
prepared to respect, is not there to keep them in line" (Anderson 1993: 85). The mere
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
absence of mature adult males may be detrimental to communities because older men can
provide the traditional male authority necessary to supplement the private networks of
community women8
Indeed, empirical research demonstrates that female ties and informal controls,
though important in controlling violence, operate at a diminished capacity in the absence
of male networks that serve to augrnent/supplement female authority (Rose and Clear
1998). Rountree and Warner (1999) find that although the level of neighborhood family
disruption does not affect the social ties of men or women, female ties are less effective
in controlling violence where males are absent. Conversely, female efforts to exercise
informal social control are most effective in communities where female-headed
households are minimal. As Anderson and Wilson propose, the physical presence of
older males in a community may provide the necessary supervision and guardianship
structures necessary to control violence. However, Rountree and Warner are not directly
measuring capacity for guardianship; rather they examine social ties (i.e., borrowing
from, dining with, and helping a neighbor) that are thought to augment a community's
capacity to exercise guardianship through a "network of collective control" (Felson 1986:
124; Thrasher 1963; Reiss 1986). In addition, their research suggests that males bring
"something" to the table other than their local networks because male ties did not
Even in areas where the adult males also engage in criminal activities and violence themselves, there is no evidence to suggest that they cannot or will not act as shepherds. As Venkatesh's (1997) research demonstrates, adult gang members involved in criminal activities also contributed to the well-being of the community by acting in supervisory capacities over youth and as protectordescorts. Furthermore, even criminal men can enhance local networks and ties thought to be vital for successful informal social control efforts (Rose and Clear 1998; Rountree and Warner 1999; Wilson 1987).
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
contribute to diminished levels of violence. One possibility is that males bring monetary
resources and other forms of social capital to communities that diminish the
incentive/motivation for violence.
Thus, the presence of old heads, particularly in localities with high father absence,
are expected to reduce levels of violence via their roles as role models, authority figures,
and providers. While Anderson suggests that old heads sometimes perform the same
functions for young women in the community as well, the majority of old head-young
adult relationships are between males. And, the focus of old heads is clearly on assisting
youth in their transition to manhood (Anderson 1990:69). Therefore, one might expect
that the crime reducing effect of old heads might be greater on male levels of violence9.
One caveat is that Anderson laments the gradual disappearance of the institution
of old heads that used to be so prominent in communities. He argues that, due to the loss
of legitimate employment opportunities in many communities, traditional old heads are
losing their prestige, authority, and credibility as role models (Anderson 1990:72).
Younger, "new" old heads are supplanting the older, responsible men as role models for
young men in the community to emulate. These younger men are products of the street
gang and indifferent to the law. Rather than gainful employment, they emphasize the
value of making a quick buck and material possessions. Further, some of the old heads
that used to be pillars of the community are resigning from their roles and may even
renounce what they once stood for. One old head that Anderson speaks with implicates
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
other old heads: "There's a lot o' young men doing the dope thing. They sell it, get high
on it.. .And I'll tell you something, as quiet as it's kept [between you and me]. There are
some old heads out here selling that dope" (1990: 73). Thus, there is some reason to
suspect that, rather than reducing violence and crime, old heads may not have the same
effect they once did and, perhaps, may even be contributing to community violence.
Community Caregivers: Grandmothers
While male capital is hypothesized to have effects mainly on male violence rates,
female old heads may have a more universal effect on crime prevention: "If old heads
were important for boys, they were also important for girls, in similar ways. The female
old heads were seen as mature and wise figures in the community, not only by women
and girls, but also by many young men" (Anderson 1990:73). Female old heads, often
referred to as othermothers, are an important source of social control and organization for
a community (Collins 1986), but they also play a critical role in offering social support
and coping mechanisms to (single) parents under stress. Typically, they are
grandmothers, usually over the age of 40 (James 1993). These women, through bonds of
kinship and networks of friendship, play an important role in communities, giving advice,
doling out discipline, and, sometimes, acting as community organizers (Anderson 1990:
73).
Grandmothers are an important source of social sanction and social control within
communities. For example, take this statement from a female old head: "When you see
Old heads might impact female levels of violence indirectly via the violence reducing effect on males -9
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
any child out there doing wrong.. .you corrects him.. .I don't care what I'm doing, if I see
somebodyfighting, arguing, or whatever.. .They come through here, get in afight, I get
out there and stop it" (Anderson 1990:74, my emphasis). In this example, the
othermother played a direct role in the prevention of violence. In a more indirect manner,
grandmothers act as role models by being upstanding citizens, providing supervision, and
offering anti-crime and anti-violence advice. Additionally, via her role as community
othermother, these grandmothers collectively organize, making up a core group in the
community, in an effort to bring more resources and other forms of social capital to their
community (Anderson 1990).
Clearly, these older women play a central role in communities; further, their role
may be enhanced where fathers are largely nonexistent, though Collins (2000) argues for
the centrality of women despite the absence of husbands and fathers (i.e., an additive
rather than an interactive effect). Taking on the role of "extra parent" or "surrogate
mother," grandmothers are supportive of the family institution and offer an outlet for
overburdened parents. Parents come to rely on grandmothers for emotional support,
particularly where violence and incarceration has taken its toll on communities (Hunter
1997). She might act as a third party who strengthens the relationship between parent
and child or she might step in to assist strained bloodmothers who lack the preparation or
desire for motherhood (Collins 2000: 180). Grandmothers serve a critical function in
communities, from taking on child-care responsibilities (Stack 1974) to sitting watch on
much of female violence is in response to male aggression or co-offending with male partners.
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
porches and intervening when necessary (Anderson 1990:4) to alleviating the stress and
strain of women and mothers attempting to fulfill many social roles.
Empirical results at the individual level, though, have been somewhat mixed in
terms of the violence-prevention benefits of grandmothers as caretakers''. At the
individual-level, some researchers have identified distinctly negative effects of co-
resident grandparenting (McLanahan and Sandefur 1994; East and Felice 1996; Under
and Cooley 1992). For example, McLanahan and Sandefur (1 994) linked co-residence
with grandparents to an increased risk of dropping out of school - a likely precursor to
delinquency andlor violence. Others find that these family forms lead to a greater
likelihood of engaging in problem behaviors (Dornbusch et a1 1 985), such as earlier
initiation of substance use (Flewelling and Bauman 1990).
Widespread co-resident grandparenting may be an indicator of social
disorganization or reduced social capital. For example, some of the leading causes of
multigenerational family formation, including unemployment, incarceration, and
substance abuse (Burton 1992; Burnette 1997; Minkler 1999), are indicators of social
disorganization. Indeed, Pebley and Rudkin (1999) suggest that grandparents step in
when parent(s) are not able to provide supervision and guidance, a lack of informal social
control that is typical of socially disorganized areas. The relevance to community-level
' O We are unaware of any research that examines whether areas rich in community caregiving and grandparenting are better able to resist violence, or crime of any type. Therefore, it is necessary to draw on individual-level empirical research to speculate about possible outcomes at the aggregate level. However, we recognize that micro- and macro- level processes likely differ and we understand that outcomes identified at the aggregate-level are not necessarily generalizeable to the individual-level.
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
violence is that areas where co-resident grandparenting is more prevalent are more likely
to have trouble resisting violence due to factors related to social disorganization (Stark
1974). Further, this type of family form is more typically poor and, with time-constraints
faced by community members due to serving multiple roles, the building of social capital
via community groups and liaisons with outside government officials is compromised.
Therefore, widespread co-resident grandparenting may be indicative of particularly acute
social disorganization and lack of social capital and, therefore, may not contribute to
violence prevention, regardless of level of father absence.
On the other hand, some individual-level research identifies positive effects of
multigenerational families and grandmothers as caretakers (Aquilino 1996; Entwisle and
Alexander 1996; Leadbeater and Bishop 1994; Pope et a1 1993). For example, living in a
multigenerational family has been found to lead to positive outcomes such as higher
educational achievement and attainment or prevention of truancy and/or dropping out
(Aquilino 1996; Entwisle and Alexander 1996). DeLeire and Kalil(2002) found that
youth living in multigenerational families had outcomes related to delinquency (e.g.,
substance use) at least as good or better than two-parent families, suggesting that
grandmothers can, indeed, overcome individual-level deficits due to absent fathers.
This shared pattern of childrearing may minimize the risks associated with
poverty, unemployment, and poor physical or mental health (DeLeire and Kalil2002). In
other words, the presence of invested grandparents in a community may mitigate some of
the negative effects of father absence and structural disadvantage on violence. To explain
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
these positive outcomes, it may be that, rather than selection effects whereby the "worst
parents" draw on grandmothers as care-givers, it is actually the "best" parent(s) that
choose to have grandmothers play crucial roles in child-rearing. Additionally, collective
child-rearing provides a set of economic resources that might lessen negative outcomes.
Greater economic stability and multiple sources of income provided by this type of living
arrangement may stem violence. Lastly, communities benefit from the additional
supervision and attention of grandparents or relatives who provide care. This type of
collective caregiving may prove especially beneficial where father absence is high.
Though it is unlikely that grandparents can be perfect substitutes for biological fathers, it
may be that the benefit of grandmothers is particularly valuable in high father absence
localities. Given the demographic characteristics of caregiving grandparents -younger
and in better health physically and mentally compared to their counterparts - these types
of grandparents can be an important resource to communities in that they not only fill
child care functions, but, as Anderson and others suggest, they act as role models, sources
of social support, supervisors, disciplinarians, and stress-relievers who have the ability to
quell violence.
In summary, the presence of male capital and caregiving grandparents in a
community may have a direct negative impact on violence, though there are some reasons
to expect the opposite as well. But, controlling for factors related to social
disorganization, co-resident grandparenting and male absence should have a direct,
negative effect on gender-disaggregated rates of violence. Additionally, if the presence
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
of community caregivers and old heads can overcome deficits related to father absence,
the significance of father absence should disappear. Further, the benefits of
othermothering and male old heads are likely to be greater where father absence is more
prevalent. In terms of gender differences in violent outcomes, unlike male old heads, the
beneficial impact of community grandmothers is expected to be gender-neutral in nature,
as both young women and men draw upon the services of grandmothers. The gendered
nature of the father absence-violence link is discussed in more depth below.
Gender and the Father Absence-Violence Relationship
Prior to the 1970s, research on female crime gave inordinate attention to family
disruption as a source of female criminality (e.g., Cowie, Cowie, and Slater 1968; Elliot
and Voss 1974; Stinchcombe 1964; Schwarfman and Clark 1967) based on the
assumption that the family would have a more pervasive impact on female behavior than
male behavior due to the presumed salience of the private sphere in women's lives.
Current individual-level research on gender differences in the effect of father absence on
deviance is mixed (McClanahan and Sandefur 1994), with some research suggesting a
greater impact on females (especially dropping out of school and early childbirth) and
other research pointing to a larger impact on males (particularly for "idleness" and
fighting). No research to this point has explored whether there are gender differences, at
the aggregate level, in the effect of father absence on male and female rates of violence.' '
I I It is important to note that aggregate-level research explores how the context of father absence impacts a community's level of violence while individual-level research explores how one's family type affects an individual's likelihood of violence.
This document is a research report submitted to the U.S. Department of Justice. This report has not been published by the Department. Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice.
Though macro-level community research has burgeoned over the past decade, a
focus on gender as an analytical variable has remained virtually absent (but see
Steffensmeier and Haynie 2000a, 2000b; Steffensmeier and Stre